The MedTech AI Visibility Index 2026 measures how consistently seventy-five medical device, diagnostics and health tech companies are named and cited when clinicians and commercial buyers ask AI engines about their categories.
The full methodology, including the query wording, the run dates, the company list, and the variance between the two passes, is published alongside these results. A ranking with a hidden method is a marketing asset. This is intended to be checked.
Why this needed measuring
A growing share of category evaluation now happens before any human contact. A clinician asks an engine about a procedure, receives a composed answer that names some companies and not others, and forms a position that no commercial team will ever see recorded.
No published benchmark existed for how medical device companies perform at that layer. Companies have been buying AI visibility services without any comparative reference for what good looks like in this category, and agencies have been selling scores with no external validation.
What was measured
The sample
Seventy-five companies with a commercial product in market, stratified into three tiers of twenty-five: large cap, mid-market, and growth stage. Comparison within a tier is what makes a ranking meaningful, because a growth-stage company and a multinational are not competing for the same citation position.
The full list is published. A benchmark whose sample is concealed cannot be replicated and should not be trusted, including this one.
Four engines, three query types, two runs
Each company was assessed across ChatGPT, Perplexity, Gemini, and Google AI Overviews, on three query types, run twice on separate days to control for variance.
• Category query: the best device or system for a named indication
• Entity query: who the company is
• Procedure query: what a clinician should know before adopting the relevant procedure
Three scores
Whether the company is named at all. Whether it is named as the source of a claim rather than mentioned in passing. And whether the description returned matches the company’s own canonical description.
The third measure is the entity consistency score, and it is the most original element of the Visibility Index. It captures something the other two cannot: whether the engine actually knows what the company is, as distinct from whether it has heard of it.
The entity layer: ten sources per company
For each company, ten sources were checked for name and description consistency: the company website, the FDA 510(k) or PMA database entry, ClinicalTrials.gov sponsor records, PubMed author affiliations, Wikipedia and Wikidata, a business registry, the LinkedIn company page, SEC filings where applicable, the three leading trade publications in each category, and the company’s own press releases across three years.
The headline findings
REPLACE THIS SECTION. Three findings, each one paragraph, drawn from the actual data. Structure each as: the finding stated plainly, the number that supports it, and the implication for a commercial leader. At least one finding should be something that surprised you, because that is the one that gets quoted.
Finding one. [Replace with the strongest headline result. Historically, the most quotable finding in benchmarks of this kind concerns how few organizations clear the entity consistency floor.]
Finding two. [Replace. Consider the relationship between company size and citation share, which is the assumption most readers will arrive with.]
Finding three. [Replace. Consider the variance between engines, which determines whether a single-engine measurement is defensible at all.]
How to read your own position
Three responses are available depending on where a company landed, and they are not interchangeable.
If you are in the top quartile
Find out why, precisely, before you assume it is durable. In most cases a small number of structural assets are doing the work: one well-structured review paper, one institutional page, one clean entity record. Document what they are, because the position is more fragile than it looks and frequently depends on an individual who could move institution.
If you are in the middle
The gap is almost always the entity layer rather than the content layer. Check name consistency across the ten sources before commissioning anything. Content published under a fragmented entity accrues to nobody, so a content investment made before the entity fix produces a smaller return than the same money spent in the other order.
If you are in the bottom quartile
This is a fixable position and the fix is cheaper than most companies assume. Reconciling names across ten sources is administrative work. Adding schema is a short technical task. Neither requires a content budget or an agency.
What the Visibility Index does not measure
Three limitations worth stating plainly, because a benchmark that does not name its own boundaries invites the reader to assume it has none.
It does not measure revenue impact. Citation share is an input to commercial performance, not a proxy for it, and anyone claiming a direct conversion rate between the two is extrapolating beyond available data.
It does not measure durability. This is a point-in-time measurement taken across two runs in a single window. How fast these positions move is not yet known, and the 2027 edition will be the first opportunity to say anything about it.
It does not account for regional variation. Queries were run in English against United States market context. A company with a strong European position may score differently than its overall commercial position would suggest.
If your company is not in the Visibility Index
Run the three queries yourself before dismissing the finding. Category query, entity query, procedure query. Two engines minimum, twice each, on separate days.
Score three things. Are you named. Are you named as the source of a claim. Does the description match your own canonical description.
Twenty minutes. It will tell you more about your commercial position going into 2027 than your last brand tracker did, and unlike a brand tracker it costs nothing.
Methodology, corrections, and the 2027 edition
The complete methodology is published at the methodology page, including every query as worded, the run dates, the scoring rubric, and the full company list.
Corrections will be published rather than made silently. If a company believes it has been scored incorrectly, the method is there to be checked against, and any resulting change will appear in the monthly Decision Record with the reasoning attached.
The 2027 edition will run in the same window with the same core method, so that year-on-year comparison is possible. Any change to the method will be published before the data collection rather than after the results.
Frequently Asked Questions
What is the MedTech AI Visibility Index?
A benchmark measuring how consistently seventy-five medical device, diagnostics and healthtech companies are named and cited when buyers ask AI engines about their categories. It assesses four engines across three query types, with an accompanying ten-source entity consistency audit for each company.
How were companies selected for the Visibility Index?
Seventy-five companies with a commercial product in market, stratified into three tiers of twenty-five covering large cap, mid-market and growth stage. Comparison within a tier is what makes the ranking meaningful. The full list is published alongside the results.
Which AI engines were tested?
ChatGPT, Perplexity, Gemini and Google AI Overviews. Each query was run twice on separate days to control for variance between runs, and the variance is reported rather than averaged away.
What is an entity consistency score?
A measure of whether an engine’s description of a company matches that company’s own canonical description, checked against ten sources including regulatory filings, trial registries and author affiliations. It captures whether an engine knows what a company is, as distinct from whether it has heard of it.
Does citation share predict revenue?
No. Citation share is an input to commercial performance rather than a proxy for it. The Visibility Index does not measure revenue impact and any claimed conversion rate between the two extrapolates beyond available data.
My company is not in the Index. How do I measure my own position?
Run three queries, a category query, an entity query and a procedure query, on at least two engines, twice each on separate days. Score whether you are named, whether you are named as the source of a claim, and whether the description matches your canonical description.
How can I check or dispute a score?
The full methodology, including query wording, run dates and scoring rubric, is published. Corrections are published rather than made silently, and any resulting change appears in the monthly Decision Record with the reasoning attached.
Will the Visibility Index be repeated?
Yes, annually in the same window with the same core method so year-on-year comparison is possible. Any methodological change will be published before data collection rather than after results.
EXTERNAL CITATIONS
• FDA 510(k) searchable database
• PubMed